Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Dec 1.
Published in final edited form as: Toxicol Appl Pharmacol. 2009 Aug 28;241(2):221–229. doi: 10.1016/j.taap.2009.08.019

Arsenicals Produce Stable Progressive Changes in DNA Methylation Patterns that are Linked to Malignant Transformation of Immortalized Urothelial Cells

Taylor J Jensen 1,2, Petr Novak 2,3, Shawn M Wnek 1, A Jay Gandolfi 1, Bernard W Futscher 1,2
PMCID: PMC2783766  NIHMSID: NIHMS143770  PMID: 19716837

Abstract

Aberrant DNA methylation participates in carcinogenesis and is a molecular hallmark of a tumor cell. Tumor cells generally exhibit a redistribution of DNA methylation resulting in global hypomethylation with regional hypermethylation; however, the speed in which these changes emerge has not been fully elucidated and may depend on the temporal location of the cell in the path from normal, finite lifespan to malignant transformation. We used a model of arsenical-induced malignant transformation of immortalized human urothelial cells and DNA methylation microarrays to examine the extent and temporal nature of changes in DNA methylation that occur during the transition from immortal to malignantly transformed. Our data presented herein suggest that during arsenical-induced malignant transformation, aberrant DNA methylation occurs non-randomly, progresses gradually at hundreds of gene promoters, alters expression of the associated gene, and these changes are coincident with the acquisition of malignant properties, such as anchorage independent growth and tumor formation in immunocompromised mice. The DNA methylation changes appear stable, since malignantly transformed cells removed from the transforming arsenical exhibited no reversion in DNA methylation levels, associated gene expression, or malignant phenotype. These data suggest that arsenicals act as epimutagens and directly link their ability to induce malignant transformation to their actions on the epigenome.

Keywords: DNA Methylation, Epigenetic, Arsenic, Histone acetylation, Bladder

Introduction

Aberrant DNA methylation occurs in nearly all tumor types and likely plays a causal role in carcinogenesis (Jones and Baylin, 2007). Relative to their normal counterparts, tumor cells generally exhibit a redistribution of DNA methylation, resulting in a pattern of global hypomethylation with regional hypermethylation (Riggs and Jones, 1983; Jones and Baylin, 2007). The loss of DNA methylation during carcinogenesis occurs primarily in repetitive elements and intragenic regions, which can result in chromosome instability and an increased mutation rate (Lengauer et al., 1997; Chen et al., 1998; Shann et al., 2008). DNA hypermethylation is typically associated with the promoter regions of genes or clusters of genes (Feinberg and Vogelstein, 1983; Goelz et al., 1985; Costello et al., 2000; Frigola et al., 2006; Novak et al., 2006; Novak et al., 2008). The DNA hypermethylation of gene promoters has been mechanistically linked to the alteration of other epigenetic modifications, such as the reduced histone acetylation, and these alterations lead to transcriptional silencing (Jones et al., 1998; Nan et al., 1998). Additionally, the changes in DNA methylation that occur during tumorigenesis are stable and may be further propagated within an individual tumor in vivo as well as in cell lines derived from these tumors (Markl et al., 2001). Recent studies have shown that the number of differentially methylated regions in an individual tumor is estimated to be in the hundreds or low thousands (Markl et al., 2001; Jones and Baylin, 2007; Novak et al., 2008; Irizarry et al., 2009).

While disruption of normal DNA methylation patterns is a known contributor to cancer, the manner in which these lesions accumulate during carcinogenesis is not completely understood and may depend upon their location in the process of malignant transformation, for example such as after the acquisition of the immortal phenotype. Early in tumorigenesis, epithelial cells go through distinct proliferative barriers that must be overcome prior to reaching immortality (Brenner et al., 1998; Romanov et al., 2001; Stampfer and Yaswen, 2003; Garbe et al., 2007). Multiple epigenetic aberrations are acquired in a precipitous fashion coincident with overcoming these barriers. Step-wise methylation changes are thus characteristic of pre-immortal stages of tumorigenesis (Novak et al., 2009). On the other hand, during the malignant transformation of an immortalized, non-tumorigenic cell or the progression of an initiated tumor cell, DNA methylation may proceed in a more gradual, progressive manner (Salem et al., 2000; Jiang et al., 2008; Watts et al., 2008).

As a model of post-immortalization epigenetic dysfunction, we used an immortalized, non-tumorigenic human urothelial cell line (UROtsa) and its malignantly transformed variants, created by independent exposures to either arsenite [As (III)] or monomethylarsonous acid [MMA (III)] (Petzoldt et al., 1995; Rossi et al., 2001; Sens et al., 2004; Bredfeldt et al., 2006). With increasing time of arsenical exposure, these cells exhibit a progressively increasing malignant phenotype. Exposure to MMA (III) for 12 weeks (URO-MSC12) results in cells with an increased growth rate. Those exposed to arsenicals for 24 weeks (URO-MSC24) grow in an anchorage independent fashion, and those exposed to either MMA (III) (URO-MSC52) or As (III) (URO-ASSC) for 52 weeks form tumors when injected into immunocompromised mice (Sens et al., 2004; Bredfeldt et al., 2006). Earlier studies have demonstrated that exposure to arsenicals results in changes in DNA methylation and histone tail modifications globally as well as in select promoter regions; however, studies have yet to examine DNA methylation changes in a genome-wide and temporal manner (Zhao et al., 1997; Jensen et al., 2008; Zhou et al., 2008; Jensen et al., 2009).

In this study, we acquired temporal snapshots of DNA methylation patterns during the arsenical-induced malignant transformation of the immortalized UROtsa cells. These DNA methylation microarray experiments detected hypomethylation of repetitive elements such as satellite sequences, in support of earlier observations (Zhao et al., 1997). In addition, these experiments also revealed changes in DNA methylation at target gene promoters that progress gradually during the process of malignant transformation, and manifest as both hyper- and hypomethylated promoter regions. Aberrant DNA methylation of target promoters was confirmed using MALDI-TOF mass spectrometry (MassARRAY) and was linked to changes in the expression of the associated gene, as assessed by quantitative real time RT-PCR. Long-term removal of arsenicals did not reverse these DNA methylation changes, gene expression, or the malignant phenotype. These data indicate that epimutagens, such as arsenicals, lead to changes in DNA methylation that occur in a gradual, progressive, and stable fashion in immortalized cells, and that these epigenetic changes are coincident with the acquisition of an increasingly malignant phenotype.

Materials and Methods

Cell Culture

UROtsa, URO-ASSC, and URO-CDSC cells were a kind gift from Drs. Donald and Maryann Sens. The UROtsa cell line was created from the urothelial cells of a 12-year-old female donor and were immortalized using a SV40 large-T antigen construct (Petzoldt et al., 1995; Rossi et al., 2001). URO-ASSC cells were created through continuous exposure of UROtsa to 1 μM As (I II) (Sens et al., 2004). URO-MSC cell lines were created through continuous exposure of UROtsa cells to 50 nM MMA (III) as previously described (Bredfeldt et al., 2006). Each of the UROtsa cell lines were cultured as previously described (Novak et al., 2008). Similar to the arsenical selection, URO-CDSC cells were created by continuous exposure of UROtsa to 1 μM cadmium for one year (Sens et al., 2004).

Nucleic Acid Isolation

Total RNA was isolated using the RNeasy Mini Kit, and genomic DNA was isolated using the DNeasy Blood and Tissue Kit according to manufacturer's protocol (Qiagen, Valencia, CA) (Qiagen, Valencia, CA). The quantity and relative quality of each sample was assessed using absorbance at 260 nm using the NanoDrop 1000 Spectrophotometer (NanoDrop, Wilmington, DE).

MeDIP Coupled to Human Promoter Microarray Analysis

Primers for the 13,000 human promoter microarray probes were obtained from the Whitehead Institute (Odom et al., 2004). Each primer pair was designed to span approximately 750 bp upstream and 250 bp downstream of the transcription start site of these genes based upon the April, 2001 build of the NCBI database (Odom et al., 2004). Methylcytosine Immunoprecipitations (MeDIP) were performed as previously described (Novak et al., 2008). Equal amounts of input and immunoprecipitated DNA were amplified using the BioPrime Array CGH Genomic Labeling Module (Invitrogen, Carlsbad, CA) according to modified manufacturer's protocol. Two rounds of amplification were performed, during the second of which input DNA was labeled with Cy-5 and immunoprecipitated DNA was labeled with Cy-3. Labeled DNA was quantified to ensure proper amplification and incorporation of cy-labeled dUTP using the microarray function of the NanoDrop 1000 Spectrophotometer (NanoDrop, Wilmington, DE). Resultant products were co-hybridized to the human promoter microarray as previously described (Novak et al., 2008).

MassARRAY DNA methylation Analysis

Sodium bisulfite (NaBS) treated genomic DNA was prepared according to manufacturer's instructions (Zymo Research, Orange, CA). NaBS treated DNA (5 ng) was seeded into a region specific PCR reaction incorporating a T7 RNA polymerase sequence as described by the manufacturer (Sequenom, San Diego, CA). Resultant PCR product was then subjected to in vitro transcription and RNase A cleavage using the MassCLEAVE T-only kit, spotted onto a Spectro CHIP array, and analyzed using the MassARRAY Compact System MALDITOF mass spectrometer (Sequenom, San Diego, CA). Each NaBS treated DNA sample was processed in two independent experiments. Data were analyzed using EpiTyper software (Sequenom, San Diego, CA) as described (Coolen et al., 2007; Novak et al., 2008). Primers were designed using Epidesigner software (http://www.epidesigner.com) to amplify bisulfite-modified DNA and analyze CpG sites contained within the probe sequence present on the human promoter microarray. Primer sequences are provided in Supplemental Table 1 and genomic locations of the regions analyzed are provided in Supplemental Figure 1.

CpG Island Microarray

Six thousand eight-hundred clones from a human CpG island library were arrayed, DNA sequence validated, and prepared for microarray as described (Cross et al., 1994; Nouzova et al., 2004). DNA methylation analysis using McrBC digestion coupled to CpG island microarray hybridization was performed as previously described (Cross et al., 1994; Nouzova et al., 2004). Briefly, genomic DNA was cut by MseI (New England Biolabs, Beverly, MA), and then a catch-linker was ligated to the MseI fragments. These fragments were then cut with a methylation-sensitive restriction enzyme, McrBC (New England Biolabs). “Mock-cut” samples were samples exposed to the same conditions and reagents of the digested samples, except no GTP was added to drive the restriction digest. Twenty nanograms of the mock-cut or uncut samples and 20 ng of McrBC-cut MseI-linked genomic DNA was amplified by PCR. The PCR products were purified with the QIAquick PCR purification kit (Qiagen). Fluorescent Cy3 or Cy5 dye was incorporated into the PCR product using the BioPrime DNA labeling system (Invitrogen, Carlsbad, CA). After the labeling, the cut and mock reactions were mixed together and again cleaned with the QIAquick PCR purification kit (Qiagen). After purification, the labeled target was lyophilized to dryness and resuspended in 10 μl of hybridization buffer [2x SSC, 0.1% SDS, 100 ng/μl Cot1 DNA, 100 ng/μl oligo(dA)], denatured by boiling for 10 min, and added to a processed array slide. A coverslip (22 × 22 mm) was applied, and the array was placed in a hybridization chamber (GeneMachines) at 62°C for 4 to 8 h. Following hybridization, slides were washed by placing them into 50-ml conical tubes containing 2x SSC, 0.1% SDS for 5 min, 0.06x SSC, 0.1% SDS for 5 min, and 0.06x SSC for 2 min, all at room temperature. Slides were scanned for Cy3 and Cy5 fluorescence using an Axon GenePix 4000 microarray reader (Axon Instruments, Inc., Foster City, CA).

Real-Time RT-PCR

Gene expression was measured using quantitative real-time RTPCR as previously described (Jensen et al., 2008). Roche Universal Probe Library information and primer sequences are provided in Supplemental Table 1.

Clonal Sodium Bisulfite Sequencing

Bisulfite sequencing was performed as previously described (Novak et al., 2006) using the method described by Clark, et al (Clark et al., 1994).

Data Analysis

All microarray data were processed in R programming environment (R, 2007). For normalization of all data, the Linear Models for Microarray Data (Limma) package was used (Smyth, 2005). Differentially methylated elements were identified using statistical approaches previously described (Smyth, 2004). To control for false discovery rate, a multiple testing correction was performed according to methods described by Benjamini and Hochberg (Benjamini and Hochberg, 1995). Regions were considered statistically significant if the adjusted p-value was p<0.05.

Gene Ontology Testing

Data from the human promoter microarrays were analyzed based upon gene ontology (GO) terms. GO terms over-representation testing was performed using GOstats package (Falcon and Gentleman, 2007). The overlapping probabilities of DMR sets were calculated using a hypergeometric test (Fury et al., 2006).

Tumor Growth in SCID Mice

A severe combined immunodeficiency (SCID) mouse colony was developed at the University of Arizona using original SCID (C.B-17/IcrACCscid) obtained from Taconic (Germantown, New York). The mice were housed in microisolator cages (Allentown Caging Equipment Company, Allentown, NJ) and maintained under specific pathogen-free conditions. Sentinel mice were screened monthly for mycoplasma, mouse hepatitis virus, pinworms, and Sendai virus via ELISA. Cells (10 × 106) from each cell line (UROtsa, URO-MSC52, and URO-MSC52 + 6mo) were injected subcutaneously. Subcutaneous tumors were measured twice a week for tumor volume estimation (mm3) in accordance with the formula (a2 × b / 2) where a is the smallest diameter and b is the largest diameter. All procedures were performed in accordance with approved protocols of the University of Arizona Institutional Animal Care and Use Committee.

Results

Aberrant DNA methylation of specific loci during Arsenical-Induced Malignant Transformation

Previous studies have reported the derivation of malignant cell line variants from immortalized urothelial cells (UROtsa) following exposure to different environmental toxicants. Table 1 describes characteristics of these variant cell lines, which provide unique models with which to examine the molecular mechanisms that underlie malignant transformation. To assess the DNA methylation patterns in our model of malignant transformation, we utilitized two complimentary techniques coupled to two different microarray platforms. First, DNA from UROtsa, URO-ASSC, and the URO-MSC cell lines was analyzed by 5-methylcytosine immunoprecipitation (MeDIP) coupled to a 13,000 element human promoter microarray (Odom et al., 2004; Jensen et al., 2008). These microarrays provided data about methylation changes localized within promoter regions and therefore candidate genes whose expression may be directly affected. To obtain information about methylation changes occurring in CpG rich regions apart from gene promoters as well as high copy repetitive elements, we also analyzed DNA methylation using an approach where DNA is digested with the methylation-sensitive restriction enzyme McrBC, amplified, and hybridized to a 6000 element CpG island microarray (Cross et al., 1994; Nouzova et al., 2004).

Table 1. List of cell lines used in this study.

Cell line name, the treatment metal, concentration (exposure), and duration of treatment for each cell line are shown. In addition, the phenotypic properties of each cell line including increased growth rate relative to UROtsa (Hyperproliferation), anchorage independent growth (AIG), and ability of each cell line to form tumors when injected subcutaneously in to immunocompromised mice is described. Reference (Ref) describes previous publications describing part or all of the information presented for a given cell line. NA=not applicable; ND=not determined.

Cell Line Treatment Exposure Duration Hyper-
proliferation		 AIG Tumors
in Mice		 Reference
UROtsa None None None NA NO NO Rossi et al., 2001;
Bredfeldt et al., 2006
URO-MSC12 MMA (III) 50 nM 12 wk YES NO NO Bredfeldt et al., 2006
URO-MSC24 MMA (III) 50 nM 24 wk YES YES NO Bredfeldt et al., 2006
URO-MSC36 MMA (III) 50 nM 36 wk YES YES ND Bredfeldt et al., 2006
URO-MSC52 MMA (III) 50 nM 52 wk YES YES YES Bredfeldt et al., 2006
URO-MSC24
+ 3mo 		MMA (III) 50 nM 24 wk ND YES ND Bredfeldt et al., 2006;
Jensen et al., 2009
URO-MSC24
+ 6mo 		MMA (III) 50 nM 24 wk ND YES ND Bredfeldt et al., 2006;
Jensen et al., 2009
URO-MSC52
+ 3mo 		MMA (III) 50 nM 52 wk ND YES YES Bredfeldt et al., 2006;
Jensen et al., 2009
URO-MSC52
+ 6mo 		MMA (III) 50 nM 52 wk ND YES YES Bredfeldt et al., 2006;
Jensen et al., 2009
URO-ASSC As (III) 1 μM 52 wk YES YES YES Sens et al., 2004
URO-CDSC Cd (II) 1 μM 52 wk YES YES YES Sens et al., 2004
Open in a new tab

The DNA methylation profile of gene promoters in each cell line was determined. MMA (III) and As (III) treated cell lines were compared to the parental UROtsa. After 52 weeks of exposure to MMA (III), the number of aberrantly methylated promoters increased by more than 400 (Figure 1A). We detected both hypermethylation and hypomethylation events with hypermethylation being the more prevalent event in gene promoters. The changes observed in promoter regions represented approximately 4% of all elements examined in URO-MSC52, demonstrating that changes in DNA methylation are a common event during arsenical-induced malignant transformation. For a list of the results for all promoter regions analyzed, refer to Supplemental Table 2.

Figure 1.

Figure 1

Open in a new tab

DNA methylation changes during arsenical-induced malignant transformation as determined by DNA methylation microarrays. (A) The number of hyper- and hypomethylated elements in each of the MSC cell lines relative to parental UROtsa cells, as detected by MeDIP coupled to human gene promoter microarray analysis. (B) Each individual line shows the progression of DNA methylation state of individual gene promoters from parental UROtsa through MSC12, MSC24, and MSC36 to MSC52 cells. Gene promoters shown are those which exhibited a statistically significant (adjusted p<0.05) change in methylation between URO-MSC52 cells and parental UROtsa. Genes that become hypermethylated (red) and hypomethylated (green) are shown. (C) As (III) and MMA (III) target common gene promoters during malignant transformation. A Venn diagram shows the number of gene promoters which were differentially methylated in a statistically significant fashion (adjusted p<0.05) in URO-MSC52 (MSC52) and URO-ASSC cell lines relative to parental UROtsa. The total number of promoters analyzed was 10350. The p-value is the probability of the overlap being due to random chance and was calculated using a hypergeometric test. (D) Single copy elements show predominantly hypermethylation while repetitive elements show significant hypomethylation in URO-MSC52 cells, as determined by McrBc-coupled CpG island microarray analysis. The y-axis values show the fraction of differentially methylated elements (adjusted p-value p<0.05) present in URO-MSC52 relative to UROtsa for each of five genomic categories. Single copy elements are designated TSS/gene (probe is located within 2000bp of transcription start site of associated gene) or no gene (probe is located greater than 2000bp from a transcription start site). Transcription start sites are designated as such according to the UCSC Genome Browser (http://genome.ucsc.edu). Ribosomal RNA elements are designated as rRNA. Repetitive elements are designated as either Alu or Satellite sequences.

We assessed how these individual promoter regions accumulated DNA methylation temporally during malignant transformation, by querying their DNA methylation state at earlier time points in the derivation of URO-MSC52. Figure 1B shows that the individual promoter regions which become hypermethylated in UROMSC52 cells began to exhibit hypermethylation early in the transformation process and these levels increased progressively through malignant transformation. A similar but opposite effect is seen in regions which become hypomethylated in URO-MSC52. Even though only five promoter (four hypermethylated, one hypomethylated) regions pass our statistical threshold in URO-MCS12, these data show changes in DNA methylation status can be detected in target promoters as early as 12 weeks after exposure to MMA (III). These data suggest that arsenic induces methylation of target promoter regions in a gradual time-dependent fashion and is linked to an increasingly malignant phenotype.

We next assessed whether the observed DNA methylation changes were targeted to specific promoter regions within the genome. URO-ASSC and URO-MSC52 cells were created independently using different arsenicals, thereby providing biological replicates of arsenical-induced malignant transformation. It was calculated that the expected number of shared aberrantly methylated promoters based on a random distribution between these two cell lines would be 6.89 - we observed 108 (Figure 1C). It is unlikely that the observed overlap between these two cell lines is due to random chance (p<2.2×10−16), based on statistical analysis using a hypergeometric test. These data suggest that chronic exposure to arsenicals may lead to non-random alterations in the DNA methylation levels of target promoter regions within immortalized human cells.

To analyze the potential biological significance of the observed alterations in methylation in URO-MSC52, we performed a GO analysis of the genes whose promoters were differentially methylated. Promoter regions were categorized using GO terms and Supplemental Table 3 shows the groups of genes which are significantly (p<0.05) enriched based upon GO classification. Hyper- and hypo-methylated elements were analyzed separately with both groups shown. Hypermethylated promoters exhibited more significantly enriched subcategories than did hypomethylated promoters. The most significant hypermethylated groups of genes when categorized based upon cellular component were those associated with the plasma membrane. Functionally, the most commonly enriched genes were those involved in receptor activity, in particular those categorized as contributing to G-protein coupled receptors. There were fewer categories which exhibited significantly enriched hypomethylated promoters. The most commonly altered were those involved in sensory perception, nucleosome assembly, and pigment metabolism. Results from this analysis provide possible cell functions and pathways that may be preferentially targeted by epigenetic control during malignant transformation.

To analyze differential DNA methylation in genomic regions located outside of gene promoter regions, UROtsa and URO-MSC52 cells were analyzed using McrBc restriction enzyme digestion coupled to CpG island microarray hybridization (Nouzova et al., 2004; Watts et al., 2008). Relative to the parental UROtsa cells, URO-MSC52 cells exhibited DNA hypermethylation more frequently in single copy regions close to transcription start sites or directly in genes while hypomethylation was more prevalent in repetitive elements such as satellite sequences (Figure 1D). Such a distribution of hyper-and hypo-methylated elements between gene-related single copy sequences and repetitive satellites is in agreement with the pattern commonly observed in many types of cancer, including bladder cancer (Salem et al., 2000; Enokida and Nakagawa, 2008), and suggests that DNA methylation changes in our UROtsa cell model recapitulate the in vivo process.

MassARRAY and Sodium Bisulfite Sequencing Confirm Microarray Data

The DNA methylation status of eleven promoter regions identified in our DNA methylation microarray studies was confirmed and extended using MassARRAY or bisulfite sequencing. The DNA methylation of KRT7, FAM83A, CYP24A1, C1QTNF6, THEM4, G0S2, NR0B1, GNA14, and EREG was measured using MassARRAY in UROtsa, URO-MSC12, URO-MSC24, URO-MSC36, URO-MSC52, and URO-ASSC. The location of the promoter microarray probe and the region analyzed using MassARRAY is shown in Supplemental Figure 1. The DNA methylation levels of each amplicon analyzed confirmed that the exposure to arsenicals leads to gradual changes in DNA methylation and the magnitude of changes grows with prolonged exposure (Figure 2; Supplemental Figure 2A). Data obtained from the second microarray platform, the CpG island microarray, were validated using clonal sodium bisulfite sequencing. Analysis of two CpG island islands associated with the genes THBD and ADAMTS5 also confirmed the methylation microarray data (Supplemental Figure 3).

Figure 2.

Figure 2

Open in a new tab

MassARRAY confirms the arsenical-mediated differential DNA methylation detected by gene promoter microarray analysis. DNA methylation levels in parental UROtsa cells and each of the malignantly transformed variants (MSC12, MSC24, and MSC36, MSC52, and ASSC cells) were assessed in six differentially methylated gene promoter regions using Sequenom MassARRAY. The values presented show the overall percent CpG methylation for each promoter region analyzed.

To test the possibility that the changes in DNA methylation could have occurred as a result of long term in vitro culture, the DNA methylation status of the nine gene promoters shown in Figure 2 and Supplemental Figure 2 were analyzed in a UROtsa cell culture at time zero and then again after six months of continuous culture. Results from the MassARRAY analysis indicate DNA methylation state did not change in any of the promoter regions due solely to continuous culture (Supplemental Figure 4).

DNA Methylation is Linked to Changes in Gene Expression and Histone H3 Acetylation

We performed quantitative real time RT-PCR on parental UROtsa cells as well as URO-ASSC and each of the URO-MSC cell lines to assess the effect of DNA methylation on the transcription of the target genes. The transcript levels of KRT7, FAM83A, C1QTNF6, THEM4, G0S2, and EREG, each confirmed to be differentially methylated, were inversely correlated with DNA methylation. Expression levels of those genes that exhibited progressive hypermethylation exhibited a loss in gene expression that mirrored this pattern (Figure 3). Taken together, these data suggest that arsenical-induced aberrant DNA methylation is linked to changes in gene expression in this model of malignant transformation.

Figure 3.

Figure 3

Open in a new tab

Decreased gene expression is linked to arsenical-mediated aberrant DNA hypermethylation of the target gene promoter. Expression levels were determined for select genes that had verified differentially methylated promoters. Quantitative real-time RT-PCR was performed on three independent samples from parental UROtsa and each of the arsenical exposed variants. The average expression value shown on the y-axis is normalized to β-Actin expression and is relative to expression levels seen in parental UROtsa cells. Error bars show the standard deviation.

The hypermethylated promoters identified in this study were integrated with the promoter histone acetylation data obtained in our previous study (Jensen et al., 2008). Normalized log2 ratios (ChIP/Input) from the human promoter microarray analysis presented herein and those published previously were plotted and correlation coefficients were calculated for each promoter region analyzed. There is a consistent inverse correlation between DNA methylation and histone H3 acetylation with negative coefficients ranging from 0.750 - 0.997 (Supplemental Figure 5). These data suggest that epigenetic remodeling of target loci in our model system involves multiple epigenetic components.

Aberrant DNA Methylation is Stable After Arsenical Removal

To assess the stability of aberrant DNA methylation in a set of target promoter regions, URO-MSC24 and URO-MSC52 cells were cultured in the absence of MMA (III) for three (URO-MSC+3 mo) and six months (URO-MSC+6 mo). DNA methylation was measured using MassARRAY technology with each sample analyzed in two independent experiments. In each of the nine promoter regions examined, DNA methylation remained elevated at levels similar to those measured in URO-MSC24 (Figure 4A, Supplemental Figure 2B) and URO-MSC52 (Figure 4B; Supplemental Figure 2C). These data indicate that the changes in DNA methylation are stable after arsenical-induced malignant transformation. Furthermore, for each of the genes analyzed, withdrawal of MMA(III) from URO-MSC24 (Figure 5A) and URO-MSC52 (Figure 5B) had no or little effect upon gene expression. These data suggest that the changes in DNA methylation observed in the malignantly transformed cells are not dependent upon the presence of the transforming arsenical and are linked to stable changes in associated gene expression.

Figure 4.

Figure 4

Open in a new tab

Altered DNA methylation patterns are stable after long term removal of arsenicals. Sequenom MassARRAY was used to analyze the DNA methylation status of each sample. (A) DNA methylation levels were determined for six differentially methylated gene promoter regions in parental UROtsa, URO-MSC24 and URO-MSC24 cells grown in the absence of MMA (III) for 3 or 6 months. (B) DNA methylation levels were determined for six differentially methylated gene promoter regions in parental UROtsa, URO-MSC52, and URO-MSC52 cells grown in the absence of MMA (III) for 3 or 6 months. The values presented show the overall percent CpG methylation for each promoter region analyzed.

Figure 5.

Figure 5

Open in a new tab

Gene expression changes are stable after long-term removal of MMA (III). (A) Gene expression levels were determined in UROtsa, URO-MSC24, and URO-MSC24 cultured in the absence of MMA (III) for 3 or 6 months using quantitative real-time RT-PCR. (B) Gene expression levels were determined in UROtsa, URO-MSC52, and URO-MSC52 cultured in the absence of MMA (III) for 3 or 6 months using quantitative real-time RT-PCR. The average expression value shown on the y-axis is normalized to β-Actin expression and is relative to expression levels seen in parental UROtsa cells.Experiments were performed on three independent samples for each cell line with the average expression value shown. Error bars show the standard deviation.

Epigenetic Stability is Linked to the Maintenance of a Tumorigenic Phenotype

As a marker of the malignant phenotype, we determined if cells removed from arsenical exposure retained the ability to form tumors as xenografts in immunocompromised mice. UROtsa, URO-MSC52, and URO-MSC52+6 mo cells were injected subcutaneously into SCID mice (n=4). Tumor growth was assessed by measuring the volume of the palpable tumor. URO-MSC52+6 mo cells formed tumors at a similar rate and size to its antecedent cell line URO-MSC52 (Figure 6). As previously demonstrated, passage matched UROtsa cells did not form tumors in mice while UROMSC52 cells did (Bredfeldt et al., 2006). These data indicate that stable epigenetic remodeling is correlated with the maintenance of a tumorigenic phenotype.

Figure 6.

Figure 6

Open in a new tab

The tumorigenic phenotype of URO-MSC52 cells persists after long-term removal of MMA (III). UROtsa, URO-MSC52, and URO-MSC52 cells cultured in the absence of MMA (III) for 6 months (MSC52+6mo) were injected into SCID mice and allowed to grow for 76 days to assess tumorigenicity. Mean tumor burden is the average tumor volume of four independent (n=4) experiments. Error bars indicate the standard error of the mean at each time point for each cell line.

Discussion

Epigenetic aberrations, including the redistribution of DNA methylation, play a participative role in carcinogenesis (Riggs and Jones, 1983; Jones and Baylin, 2007). The rate and pattern in which these epigenomic alterations emerge is still being elucidated and may vary depending on the location of a cell along its path from normality to malignancy. The link between arsenic exposure and cancer of the lung, skin, liver, and bladder has been well established; however, the mechanisms by which arsenicals participate in carcinogenesis are not completely understood (Chen et al., 1985; Marshall et al., 2007). One mechanism by which arsenicals appear to participate is through the altering the epigenetic landscape (Zhao et al., 1997; Chen et al., 2004; Marsit et al., 2006; Xie et al., 2007; Jensen et al., 2008; Jensen et al., 2009). Using an in vitro human urothelial cell line model of arsenical-induced malignant transformation, we set out to identify the gene targets of epigenetic change, the speed of the changes and their stability, and how these changes related to the emergence of malignant characteristics.

We performed DNA methylation microarrays and found that hundreds of alterations in DNA methylation occur during arsenical-induced malignant transformation. These data are consistent with previous studies demonstrating that exposure to arsenic results in the loss of global DNA methylation and the alteration of promoter region DNA methylation in a small number of regions examined (Zhao et al., 1997; Chen et al., 2004; Marsit et al., 2006; Xie et al., 2007; Jensen et al., 2008). These data thereby reinforce that there is likely an epigenetic component of arsenic-associated malignancy. The data presented herein also show that arsenical-induced aberrant DNA methylation is acquired gradually during the malignant transformation of immortalized cells. The changes in DNA methylation occur non-randomly and are linked to altered transcriptional regulation of affected genes. These changes appear stable since long-term removal of the arsenical from the malignantly transformed cells did not result in reversion of promoter region DNA methylation, associated gene expression, and a malignant phenotype. These data suggest environmentally relevant species of arsenicals can act as epimutagens and may participate in human carcinogenesis by altering the epigenetic landscape of immortalized cells.

Furthermore, aberrant DNA methylation is known to act in concert with histone tail modifications, in particular histone acetylation, to regulate gene expression. For example, members of the methyl-binding domain (MBD) class of proteins bind to methylated cytosines and recruit histone deacetylase (HDAC) enzymes, resulting in the concurrence of DNA hypermethylation, histone hypoacetylation and transcriptional silencing (Jones et al., 1998; Nan et al., 1998). Overall, our data show gene targets of arsenical exposure show strong correlations between their DNA methylation, histone acetylation, and gene expression states, suggesting that this epigenetic control pathway may participate in arsenical-mediated malignant transformation (Supplemental Figure 5).

The DNA methylation microarray analyses presented herein uncovered hundreds of aberrantly methylated genes that may warrant future study, of which the gene G0S2 (G0/G1 Switch Gene 2) may prove to be of particular interest. The CpG island promoter region associated with G0S2 is an early target of epigenetic remodeling during arsenical-induced malignant transformation with significant changes detected after 12 weeks of exposure. G0S2 exhibits promoter hypermethylation and transcriptional silencing in URO-MSC12 cells, coincident with the acquisition of hyperproliferation (Bredfeldt et al., 2006). Recent studies suggest that G0S2 likely plays a functional role in cell cycle regulation and differentiation (Zandbergen et al., 2005; Kitareewan et al., 2008). Consistent with these results, G0S2 is expressed at low levels in rapidly proliferating hepatoma cell lines when compared to growth-arrested murine liver cells, indicating a potential role in carcinogenesis as well as cell cycle regulation (Zandbergen et al., 2005). Taken together with its potential functional importance, hypermethylation of the G0S2 promoter may provide an early biomarker for the detection of pre-malignant cellular changes associated with arsenic exposure.

Since epigenetic patterns can be heritably transmitted from parent to daughter cell providing a means of cellular memory (Bird, 2002), changes to the patterns and levels of these marks by an epimutagenic toxicant may compromise the epigenetic cellular memory of subsequent generations of cells. Indeed, the changes in DNA methylation after arsenic exposure appear to be stable; therefore, they may act as a long-term detectable mark of arsenic exposure. Interestingly, these aberrations are stable even in those cells that have not yet reached complete malignant transformation, such as in URO-MSC24 cells. As such, arsenical exposure may affect normal cellular physiology before the appearance of pathological lesions, and this exposure may result in epigenetic changes that exist and can be detected as earlier markers of risk. Consistent with these findings have been epidemiological studies performed in humans that note that the increase in the relative risk of developing cancer after arsenic exposure may not occur until 20-50 years after exposure (Morales et al., 2000; Smith et al., 2006; Marshall et al., 2007; Ahlborn et al., 2008). A similar theory has been proposed that suggests that the alteration of DNA methylation at specific loci within gastric mucosae is a stable lesion that results due to exposure to heliobacter pylori (Maekita et al., 2006). Taken together, these data suggest that arsenic exposure may alter epigenetic memory prior to malignancy, and these epigenetic changes may ultimately prove useful as stable biomarkers of arsenical exposure.

These data lead us to question whether the aberrations detected in URO-ASSC and URO-MSC cells were arsenic specific or whether they would also be found in cells which were malignantly transformed using other metalloids or transition metals. Cadmium is a naturally occurring transition metal that is known to be associated with human cancer, potentially inducing carcinogenesis through epigenetic mechanisms (Takiguchi et al., 2003; Waalkes, 2003). Additionally, cadmium has been used to malignantly transform UROtsa cells, producing the URO-CDSC cell line (Sens et al., 2004). To address whether the changes found in our study were specific to arsenic or whether they may be common to multiple metals, the levels of the six transcripts described in Figure 3 were measured in URO-CDSC cells using real-time RT-PCR. Results indicate that four of the six (67%) genes examined showed transcriptional suppression in URO-CDSC similar to the levels seen in the arsenical associated cell lines (Supplemental Figure 6A). Each of these downregulated genes was associated with a hypermethylated gene promoter (Supplemental Figure 6B). These data suggest that the carcinogenicity of various metals may be linked by a common ability to remodel the epigenetic state of specific loci; future studies will explore this prediction in more detail.

Although changes in DNA methylation are evident during arsenical-induced malignant transformation, the mechanism by which these changes occur is not well understood. DNA methyltransferase (DNMT) enzymes, including the maintenance methyltransferase DNMT1, catalyze the methylation of cytosine residues within DNA. In order to determine whether the changes that observed were the result of altered expression of DNMT1, we performed quantitative real-time RT-PCR to quantify its expression (Supplemental Figure 7). The results show that DNMT1 expression is not altered during arsenical-mediated malignant transformation suggesting that other mechanisms are likely responsible for the multifaceted epigenetic changes observed during arsenical exposure. Future studies will focus on delineating the mechanisms involved in the epigenetic remodeling associated with arsenical-induced malignant transformation.

In conclusion, this study links arsenical-induced malignant transformation to the changes that the arsenicals induce in DNA methylation state. The DNA methylation changes in gene promoters emerge and progress over time, are linked to altered expression of the associated gene, and appear concurrently with the acquisition of increasingly malignant properties. Importantly, target promoter DNA methylation, associated gene expression, and the malignant phenotype were stable in the absence of the arsenical. Taken together, these studies provide early insights into the dynamics of the DNA methylation changes at specific loci that occur during arsenic-driven cancer progression as well as to establish further arsenicals as epimutagens associated with arsenic-mediated carcinogenesis.

Supplementary Material

Supplemental Figure 1

The genomic location of confirmed aberrantly methylated gene promoter regions. The top portion of each figure is taken from the UCSC genome browser (http://genome.ucsc.edu) and depicts the location of the human promoter microarray probe, the region analyzed using MassARRAY, the associated gene, and any CpG islands present in this region. The bottom portion of each figure shows the observed/expected ratios (Observed vs Expected) and GC content (Percentage) across the analyzed region. Data analysis was performed and images were created using the EMBOSS CpGPlot tool (http://www.ebi.ac.uk/Tools/emboss/cpgplot/index.html).

10
02
Supplemental Figure 2

(A) MassARRAY confirms the arsenical-mediated differential DNA methylation detected by gene promoter microarray analysis. DNA methylation levels in parental UROtsa cells and each of the malignantly transformed variants (MSC12, MSC24, and MSC36, MSC52, and ASSC cells) were assessed in three differentially methylated gene promoter regions using Sequenom MassARRAY. (B) DNA methylation levels were determined for three differentially methylated gene promoter regions in parental UROtsa, URO-MSC24 and URO-MSC24 cells grown in the absence of MMA (III) for 3 or 6 months using Sequenom MassARRAY. (C) DNA methylation levels were determined for three differentially methylated gene promoter regions in parental UROtsa, URO-MSC52, and URO-MSC52 cells grown in the absence of MMA (III) for 3 or 6 months. The values presented in each graph show the overall percent CpG methylation for each promoter region analyzed.

Supplemental Figure 3

Sodium bisulfite sequencing confirms the McrBC – CpG island microarray results. Sodium bisulfite sequencing was performed to examine the DNA methylation levels and pattern in the regions associated with THBD and ADAMTS5 identified by CpG island microarray analysis. Each column represents an individual CpG site while each row is representative of an individual sequenced clone. The methylation state at each CpG dinucleotide [methylated (■); unmethylated (□); poor sequence (■)] is shown. The methylation level (%) for the entire amplicon in each cell line is shown. Clones were sorted from low to high methylation for presentation.

Supplemental Figure 4

Long-term culture of parental UROtsa cells does not alter DNA methylation. Parental UROtsa and UROtsa cells cultured using standard conditions for six months (LT-URO) were assayed for DNA methylation in the promoter regions of nine genes using MassARRAY. Values presented show the average percent methylation for each of the promoters analyzed.

06
Supplemental Figure 6

Cadmium induces transcriptional downregulation associated with aberrant promoter methylation. (A) Quantitative real-time RT-PCR was performed on for the genes presented in Figure 3. Three independent samples from parental UROtsa and URO-CDSC were used for analysis. The average expression value shown on the y-axis is normalized to β-Actin expression and is relative to expression levels seen in parental UROtsa cells. Error bars show the standard deviation. (B) DNA methylation levels in parental UROtsa cells and URO-CDSC cells were assessed in the promoters of the four genes which were differentially expressed using Sequenom MassARRAY. The values presented show the overall percent CpG methylation for each promoter region analyzed. NA is representative of a promoter that was not analyzed.

08
Supplemental Figure 5

DNA methylation is inversely correlated with histone H3 acetylation in target gene promoter regions. Log2 ratios (ChIP/input) obtained from MeDIP and acetylated histone H3 ChIP experiments coupled to the same human promoter microarray were plotted. Data from UROtsa, URO-MSC24, URO-MSC52, and URO-ASSC were used for analysis. The correlation coefficient between DNA methylation and histone H3 acetylation for each analyzed promoter region is shown.

Acknowledgements

The authors would like to thank Drs. Donald and Maryann Sens for kindly providing the UROtsa and URO-ASSC cell lines. In addition, the authors would like to thank Dr. Kylee Eblin, Rhiannon Hardwick, and Lucina Lizarraga for assistance on various aspects of this project. The authors thank Dr. George Watts and the University of Arizona Genomics Shared Service at the Arizona Cancer Center for expertise on the microarray studies. NIEHS grant CA127989 to BF supported this research. The Genomics Shared Service is supported by NIEHS grant ES06694, NIH grant CA23074, and the UA Bio-5 Institute. Cancer Biology Training Grant CA09213 supported TJ and NIEHS Training Grant ES007091 supported SW.

Abbreviations

As (III)

Arsenite

MeDIP

methylcytosine immunoprecipitation

MMA (III)

monomethylarsonous acid

SCID

severe combined immunodeficiency

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of Interest Statement: The authors declare that there are no conflicts of interest

References

  1. Ahlborn GJ, Nelson GM, Ward WO, Knapp G, Allen JW, Ouyang M, Roop BC, Chen Y, O'Brien T, Kitchin KT, Delker DA. Dose response evaluation of gene expression profiles in the skin of K6/ODC mice exposed to sodium arsenite. Toxicol Appl Pharmacol. 2008;227:400–416. doi: 10.1016/j.taap.2007.10.029. [DOI] [PubMed] [Google Scholar]
  2. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate - a Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B-Methodological. 1995;57:289–300. [Google Scholar]
  3. Bird A. DNA methylation patterns and epigenetic memory. Genes Dev. 2002;16:6–21. doi: 10.1101/gad.947102. [DOI] [PubMed] [Google Scholar]
  4. Bredfeldt TG, Jagadish B, Eblin KE, Mash EA, Gandolfi AJ. Monomethylarsonous acid induces transformation of human bladder cells. Toxicol Appl Pharmacol. 2006;216:69–79. doi: 10.1016/j.taap.2006.04.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brenner AJ, Stampfer MR, Aldaz CM. Increased p16 expression with first senescence arrest in human mammary epithelial cells and extended growth capacity with p16 inactivation. Oncogene. 1998;17:199–205. doi: 10.1038/sj.onc.1201919. [DOI] [PubMed] [Google Scholar]
  6. Chen CJ, Chuang YC, Lin TM, Wu HY. Malignant neoplasms among residents of a blackfoot disease-endemic area in Taiwan: high-arsenic artesian well water and cancers. Cancer Res. 1985;45:5895–5899. [PubMed] [Google Scholar]
  7. Chen H, Li S, Liu J, Diwan BA, Barrett JC, Waalkes MP. Chronic inorganic arsenic exposure induces hepatic global and individual gene hypomethylation: implications for arsenic hepatocarcinogenesis. Carcinogenesis. 2004;25:1779–1786. doi: 10.1093/carcin/bgh161. [DOI] [PubMed] [Google Scholar]
  8. Chen RZ, Pettersson U, Beard C, Jackson-Grusby L, Jaenisch R. DNA hypomethylation leads to elevated mutation rates. Nature. 1998;395:89–93. doi: 10.1038/25779. [DOI] [PubMed] [Google Scholar]
  9. Clark SJ, Harrison J, Paul CL, Frommer M. High sensitivity mapping of methylated cytosines. Nucleic Acids Res. 1994;22:2990–2997. doi: 10.1093/nar/22.15.2990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Coolen MW, Statham AL, Gardiner-Garden M, Clark SJ. Genomic profiling of CpG methylation and allelic specificity using quantitative high-throughput mass spectrometry: critical evaluation and improvements. Nucleic Acids Res. 2007;35:e119. doi: 10.1093/nar/gkm662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Costello JF, Fruhwald MC, Smiraglia DJ, Rush LJ, Robertson GP, Gao X, Wright FA, Feramisco JD, Peltomaki P, Lang JC, Schuller DE, Yu L, Bloomfield CD, Caligiuri MA, Yates A, Nishikawa R, Huang HJS, Petrelli NJ, Zhang XL, O'Dorisio MS, Held WA, Cavenee WK, Plass C. Aberrant CpG-island methylation has non-random and tumour-type-specific patterns. Nature Genetics. 2000;24:132–138. doi: 10.1038/72785. [DOI] [PubMed] [Google Scholar]
  12. Cross SH, Charlton JA, Nan X, Bird AP. Purification of CpG islands using a methylated DNA binding column. Nat Genet. 1994;6:236–244. doi: 10.1038/ng0394-236. [DOI] [PubMed] [Google Scholar]
  13. Enokida H, Nakagawa M. Epigenetics in bladder cancer. Int J Clin Oncol. 2008;13:298–307. doi: 10.1007/s10147-008-0811-1. [DOI] [PubMed] [Google Scholar]
  14. Falcon S, Gentleman R. Using GOstats to test gene lists for GO term association. Bioinformatics. 2007;23:257–258. doi: 10.1093/bioinformatics/btl567. [DOI] [PubMed] [Google Scholar]
  15. Feinberg AP, Vogelstein B. Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature. 1983;301:89–92. doi: 10.1038/301089a0. [DOI] [PubMed] [Google Scholar]
  16. Frigola J, Song J, Stirzaker C, Hinshelwood RA, Peinado MA, Clark SJ. Epigenetic remodeling in colorectal cancer results in coordinate gene suppression across an entire chromosome band. Nat Genet. 2006;38:540–549. doi: 10.1038/ng1781. [DOI] [PubMed] [Google Scholar]
  17. Fury W, Batliwalla F, Gregersen PK, Li W. Overlapping probabilities of top ranking gene lists, hypergeometric distribution, and stringency of gene selection criterion. Conf Proc IEEE Eng Med Biol Soc. 2006;1:5531–5534. doi: 10.1109/IEMBS.2006.260828. [DOI] [PubMed] [Google Scholar]
  18. Garbe JC, Holst CR, Bassett E, Tlsty T, Stampfer MR. Inactivation of p53 function in cultured human mammary epithelial cells turns the telomere-length dependent senescence barrier from agonescence into crisis. Cell Cycle. 2007;6:1927–1936. doi: 10.4161/cc.6.15.4519. [DOI] [PubMed] [Google Scholar]
  19. Goelz SE, Vogelstein B, Hamilton SR, Feinberg AP. Hypomethylation of DNA from Benign and Malignant Human-Colon Neoplasms. Science. 1985;228:187–190. doi: 10.1126/science.2579435. [DOI] [PubMed] [Google Scholar]
  20. Irizarry RA, Ladd-Acosta C, Wen B, Wu Z, Montano C, Onyango P, Cui H, Gabo K, Rongione M, Webster M, Ji H, Potash JB, Sabunciyan S, Feinberg AP. The human colon cancer methylome shows similar hypo- and hypermethylation at conserved tissue-specific CpG island shores. Nat Genet. 2009 doi: 10.1038/ng.298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jensen TJ, Novak P, Eblin KE, Gandolfi AJ, Futscher BW. Epigenetic remodeling during arsenical-induced malignant transformation. Carcinogenesis. 2008;29:1500–1508. doi: 10.1093/carcin/bgn102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Jensen TJ, Wozniak RJ, Eblin KE, Wnek SM, Gandolfi AJ, Futscher BW. Epigenetic mediated transcriptional activation of WNT5A participates in arsenical-associated malignant transformation. Toxicol Appl Pharmacol. 2009;235:39–46. doi: 10.1016/j.taap.2008.10.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jiang Y, Dunbar A, Gondek LP, Mohan S, Rataul M, O'Keefe C, Sekeres M, Saunthararajah Y, Maciejewski JP. Aberrant DNA methylation is a dominant mechanism in MDS progression to AML. Blood. 2008 doi: 10.1182/blood-2008-06-163246. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Jones PA, Baylin SB. The epigenomics of cancer. Cell. 2007;128:683–692. doi: 10.1016/j.cell.2007.01.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J, Wolffe AP. Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet. 1998;19:187–191. doi: 10.1038/561. [DOI] [PubMed] [Google Scholar]
  26. Kitareewan S, Blumen S, Sekula D, Bissonnette RP, Lamph WW, Cui Q, Gallagher R, Dmitrovsky E. G0S2 is an all-trans-retinoic acid target gene. Int J Oncol. 2008;33:397–404. [PMC free article] [PubMed] [Google Scholar]
  27. Lengauer C, Kinzler KW, Vogelstein B. DNA methylation and genetic instability in colorectal cancer cells. Proc Natl Acad Sci U S A. 1997;94:2545–2550. doi: 10.1073/pnas.94.6.2545. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Maekita T, Nakazawa K, Mihara M, Nakajima T, Yanaoka K, Iguchi M, Arii K, Kaneda A, Tsukamoto T, Tatematsu M, Tamura G, Saito D, Sugimura T, Ichinose M, Ushijima T. High levels of aberrant DNA methylation in Helicobacter pylori-infected gastric mucosae and its possible association with gastric cancer risk. Clin Cancer Res. 2006;12:989–995. doi: 10.1158/1078-0432.CCR-05-2096. [DOI] [PubMed] [Google Scholar]
  29. Markl ID, Cheng J, Liang G, Shibata D, Laird PW, Jones PA. Global and gene-specific epigenetic patterns in human bladder cancer genomes are relatively stable in vivo and in vitro over time. Cancer Res. 2001;61:5875–5884. [PubMed] [Google Scholar]
  30. Marshall G, Ferreccio C, Yuan Y, Bates MN, Steinmaus C, Selvin S, Liaw J, Smith AH. Fifty-year study of lung and bladder cancer mortality in Chile related to arsenic in drinking water. J Natl Cancer Inst. 2007;99:920–928. doi: 10.1093/jnci/djm004. [DOI] [PubMed] [Google Scholar]
  31. Marsit CJ, Karagas MR, Danaee H, Liu M, Andrew A, Schned A, Nelson HH, Kelsey KT. Carcinogen exposure and gene promoter hypermethylation in bladder cancer. Carcinogenesis. 2006;27:112–116. doi: 10.1093/carcin/bgi172. [DOI] [PubMed] [Google Scholar]
  32. Morales KH, Ryan L, Kuo TL, Wu MM, Chen CJ. Risk of internal cancers from arsenic in drinking water. Environ Health Perspect. 2000;108:655–661. doi: 10.1289/ehp.00108655. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, Bird A. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature. 1998;393:386–389. doi: 10.1038/30764. [DOI] [PubMed] [Google Scholar]
  34. Nouzova M, Holtan N, Oshiro MM, Isett RB, Munoz-Rodriguez JL, List AF, Narro ML, Miller SJ, Merchant NC, Futscher BW. Epigenomic changes during leukemia cell differentiation: analysis of histone acetylation and cytosine methylation using CpG island microarrays. J Pharmacol Exp Ther. 2004;311:968–981. doi: 10.1124/jpet.104.072488. [DOI] [PubMed] [Google Scholar]
  35. Novak P, Jensen T, Oshiro MM, Watts GS, Kim CJ, Futscher BW. Agglomerative epigenetic aberrations are a common event in human breast cancer. Cancer Res. 2008;68:8616–8625. doi: 10.1158/0008-5472.CAN-08-1419. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Novak P, Jensen T, Oshiro MM, Wozniak RJ, Nouzova M, Watts GS, Klimecki WT, Kim C, Futscher BW. Epigenetic inactivation of the HOXA gene cluster in breast cancer. Cancer Res. 2006;66:10664–10670. doi: 10.1158/0008-5472.CAN-06-2761. [DOI] [PubMed] [Google Scholar]
  37. Novak P, Jensen TJ, Garbe JC, Stampfer MR, Futscher BW. Stepwise DNA methylation changes are linked to escape from defined proliferation barriers and mammary epithelial cell immortalization. Cancer Res. 2009;69:5251–5258. doi: 10.1158/0008-5472.CAN-08-4977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Odom DT, Zizlsperger N, Gordon DB, Bell GW, Rinaldi NJ, Murray HL, Volkert TL, Schreiber J, Rolfe PA, Gifford DK, Fraenkel E, Bell GI, Young RA. Control of pancreas and liver gene expression by HNF transcription factors. Science. 2004;303:1378–1381. doi: 10.1126/science.1089769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Petzoldt JL, Leigh IM, Duffy PG, Sexton C, Masters JR. Immortalisation of human urothelial cells. Urol Res. 1995;23:377–380. doi: 10.1007/BF00698738. [DOI] [PubMed] [Google Scholar]
  40. R TCD. R: A language and environment for statistical computing. R Foundation for Statistical Computing; Vienna, Austria: 2007. [Google Scholar]
  41. Riggs AD, Jones PA. 5-methylcytosine, gene regulation, and cancer. Adv Cancer Res. 1983;40:1–30. doi: 10.1016/s0065-230x(08)60678-8. [DOI] [PubMed] [Google Scholar]
  42. Romanov SR, Kozakiewicz BK, Holst CR, Stampfer MR, Haupt LM, Tlsty TD. Normal human mammary epithelial cells spontaneously escape senescence and acquire genomic changes. Nature. 2001;409:633–637. doi: 10.1038/35054579. [DOI] [PubMed] [Google Scholar]
  43. Rossi MR, Masters JR, Park S, Todd JH, Garrett SH, Sens MA, Somji S, Nath J, Sens DA. The immortalized UROtsa cell line as a potential cell culture model of human urothelium. Environ Health Perspect. 2001;109:801–808. doi: 10.1289/ehp.01109801. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Salem C, Liang G, Tsai YC, Coulter J, Knowles MA, Feng AC, Groshen S, Nichols PW, Jones PA. Progressive increases in de novo methylation of CpG islands in bladder cancer. Cancer Res. 2000;60:2473–2476. [PubMed] [Google Scholar]
  45. Sens DA, Park S, Gurel V, Sens MA, Garrett SH, Somji S. Inorganic cadmium- and arsenite-induced malignant transformation of human bladder urothelial cells. Toxicol Sci. 2004;79:56–63. doi: 10.1093/toxsci/kfh086. [DOI] [PubMed] [Google Scholar]
  46. Shann YJ, Cheng C, Chiao CH, Chen DT, Li PH, Hsu MT. Genome-wide mapping and characterization of hypomethylated sites in human tissues and breast cancer cell lines. Genome Res. 2008;18:791–801. doi: 10.1101/gr.070961.107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Smith AH, Marshall G, Yuan Y, Ferreccio C, Liaw J, von Ehrenstein O, Steinmaus C, Bates MN, Selvin S. Increased mortality from lung cancer and bronchiectasis in young adults after exposure to arsenic in utero and in early childhood. Environ Health Perspect. 2006;114:1293–1296. doi: 10.1289/ehp.8832. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Smyth GK. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004;3 doi: 10.2202/1544-6115.1027. Article3. [DOI] [PubMed] [Google Scholar]
  49. Smyth GK. Limma: linear models for microarray data. In: Gentleman R, Carey V, Dudoit S, Irizarry R, Huber W, editors. Bioinformatics and Computational Biology Solutions using R and Bioconductor. Springer; New York: 2005. pp. 397–420. [Google Scholar]
  50. Stampfer MR, Yaswen P. Human epithelial cell immortalization as a step in carcinogenesis. Cancer Lett. 2003;194:199–208. doi: 10.1016/s0304-3835(02)00707-3. [DOI] [PubMed] [Google Scholar]
  51. Takiguchi M, Achanzar WE, Qu W, Li G, Waalkes MP. Effects of cadmium on DNA-(Cytosine-5) methyltransferase activity and DNA methylation status during cadmium-induced cellular transformation. Exp Cell Res. 2003;286:355–365. doi: 10.1016/s0014-4827(03)00062-4. [DOI] [PubMed] [Google Scholar]
  52. Waalkes MP. Cadmium carcinogenesis. Mutat Res. 2003;533:107–120. doi: 10.1016/j.mrfmmm.2003.07.011. [DOI] [PubMed] [Google Scholar]
  53. Watts GS, Futscher BW, Holtan N, Degeest K, Domann FE, Rose SL. DNA methylation changes in ovarian cancer are cumulative with disease progression and identify tumor stage. BMC Med Genomics. 2008;1:47. doi: 10.1186/1755-8794-1-47. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Xie Y, Liu J, Benbrahim-Tallaa L, Ward JM, Logsdon D, Diwan BA, Waalkes MP. Aberrant DNA methylation and gene expression in livers of newborn mice transplacentally exposed to a hepatocarcinogenic dose of inorganic arsenic. Toxicology. 2007;236:7–15. doi: 10.1016/j.tox.2007.03.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zandbergen F, Mandard S, Escher P, Tan NS, Patsouris D, Jatkoe T, Rojas-Caro S, Madore S, Wahli W, Tafuri S, Muller M, Kersten S. The G0/G1 switch gene 2 is a novel PPAR target gene. Biochem J. 2005;392:313–324. doi: 10.1042/BJ20050636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhao CQ, Young MR, Diwan BA, Coogan TP, Waalkes MP. Association of arsenic-induced malignant transformation with DNA hypomethylation and aberrant gene expression. Proc Natl Acad Sci U S A. 1997;94:10907–10912. doi: 10.1073/pnas.94.20.10907. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhou X, Sun H, Ellen TP, Chen H, Costa M. Arsenite alters global histone H3 methylation. Carcinogenesis. 2008;29:1831–1836. doi: 10.1093/carcin/bgn063. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Figure 1

The genomic location of confirmed aberrantly methylated gene promoter regions. The top portion of each figure is taken from the UCSC genome browser (http://genome.ucsc.edu) and depicts the location of the human promoter microarray probe, the region analyzed using MassARRAY, the associated gene, and any CpG islands present in this region. The bottom portion of each figure shows the observed/expected ratios (Observed vs Expected) and GC content (Percentage) across the analyzed region. Data analysis was performed and images were created using the EMBOSS CpGPlot tool (http://www.ebi.ac.uk/Tools/emboss/cpgplot/index.html).

NIHMS143770-supplement-01.pdf (878.8KB, pdf)
10
NIHMS143770-supplement-10.pdf (21.1KB, pdf)
02
NIHMS143770-supplement-02.pdf (17.4KB, pdf)
Supplemental Figure 2

(A) MassARRAY confirms the arsenical-mediated differential DNA methylation detected by gene promoter microarray analysis. DNA methylation levels in parental UROtsa cells and each of the malignantly transformed variants (MSC12, MSC24, and MSC36, MSC52, and ASSC cells) were assessed in three differentially methylated gene promoter regions using Sequenom MassARRAY. (B) DNA methylation levels were determined for three differentially methylated gene promoter regions in parental UROtsa, URO-MSC24 and URO-MSC24 cells grown in the absence of MMA (III) for 3 or 6 months using Sequenom MassARRAY. (C) DNA methylation levels were determined for three differentially methylated gene promoter regions in parental UROtsa, URO-MSC52, and URO-MSC52 cells grown in the absence of MMA (III) for 3 or 6 months. The values presented in each graph show the overall percent CpG methylation for each promoter region analyzed.

NIHMS143770-supplement-03.tif (1.5MB, tif)
Supplemental Figure 3

Sodium bisulfite sequencing confirms the McrBC – CpG island microarray results. Sodium bisulfite sequencing was performed to examine the DNA methylation levels and pattern in the regions associated with THBD and ADAMTS5 identified by CpG island microarray analysis. Each column represents an individual CpG site while each row is representative of an individual sequenced clone. The methylation state at each CpG dinucleotide [methylated (■); unmethylated (□); poor sequence (■)] is shown. The methylation level (%) for the entire amplicon in each cell line is shown. Clones were sorted from low to high methylation for presentation.

NIHMS143770-supplement-04.tif (1.5MB, tif)
Supplemental Figure 4

Long-term culture of parental UROtsa cells does not alter DNA methylation. Parental UROtsa and UROtsa cells cultured using standard conditions for six months (LT-URO) were assayed for DNA methylation in the promoter regions of nine genes using MassARRAY. Values presented show the average percent methylation for each of the promoters analyzed.

NIHMS143770-supplement-05.tif (1.4MB, tif)
06
NIHMS143770-supplement-06.pdf (5.4MB, pdf)
Supplemental Figure 6

Cadmium induces transcriptional downregulation associated with aberrant promoter methylation. (A) Quantitative real-time RT-PCR was performed on for the genes presented in Figure 3. Three independent samples from parental UROtsa and URO-CDSC were used for analysis. The average expression value shown on the y-axis is normalized to β-Actin expression and is relative to expression levels seen in parental UROtsa cells. Error bars show the standard deviation. (B) DNA methylation levels in parental UROtsa cells and URO-CDSC cells were assessed in the promoters of the four genes which were differentially expressed using Sequenom MassARRAY. The values presented show the overall percent CpG methylation for each promoter region analyzed. NA is representative of a promoter that was not analyzed.

NIHMS143770-supplement-07.tif (1.4MB, tif)
08
NIHMS143770-supplement-08.pdf (62.2KB, pdf)
Supplemental Figure 5

DNA methylation is inversely correlated with histone H3 acetylation in target gene promoter regions. Log2 ratios (ChIP/input) obtained from MeDIP and acetylated histone H3 ChIP experiments coupled to the same human promoter microarray were plotted. Data from UROtsa, URO-MSC24, URO-MSC52, and URO-ASSC were used for analysis. The correlation coefficient between DNA methylation and histone H3 acetylation for each analyzed promoter region is shown.

NIHMS143770-supplement-09.pdf (18.9KB, pdf)

RESOURCES